PALAIOS, 2006, v. 21, p. 442–450
Research Article
DOI: 10.2110/palo.2005.P05-083R
QUANTIFYING TAPHONOMIC BIAS IN MOLLUSCAN DEATH ASSEMBLAGES FROM THE
UPPER CHESAPEAKE BAY: PATTERNS OF SHELL DAMAGE
ROWAN LOCKWOOD* and LAUREN A. WORK
Department of Geology, The College of William and Mary, P.O. Box 8795, Williamsburg, VA 23187, USA
e-mail: [email protected]
ABSTRACT
This study focuses on two main questions: (1) what types of shell
damage occur in the death assemblage of upper Chesapeake Bay
benthic mollusks; and (2) how does shell damage differ according to
intrinsic factors such as life habit, shell mineralogy, and shell organic
content. Extrinsic and intrinsic factors, ranging from the environment to shell composition, interact to influence the quality of fossil
preservation. Our understanding of how extrinsic factors affect shelldamage profiles has improved dramatically with the development of
taphofacies models, but the role that intrinsic factors play is still
poorly understood. Molluscan death-assemblage material was obtained via box coring, identified, and assigned taphonomic damage
states. The most common forms of shell damage were disarticulation,
fine-scale surface alteration (FSA), periostracum loss, edge modification, and fragmentation. Four patterns were documented consistently across habitat types when shell damage was examined according to life habit and shell composition. Infaunal specimens exhibit
significantly more severe damage due to internal FSA than epifaunal
specimens. Calcitic specimens experience higher levels of external encrustation than noncalcitic specimens. Specimens with high levels of
shell organics experience significantly more fragmentation and edge
modification than specimens with low levels of shell organic content.
The direction and degree to which other damage variables differ
among intrinsic categories vary according to the variable assessed or
the habitat type sampled. In the northern sites (i.e., deeper water),
all the variables that recorded differential damage according to shellorganic content recorded greater damage in specimens with high
shell organic content, suggesting that the latter may be experiencing
selective removal from the death assemblage.
INTRODUCTION
The habitats and molluscan diversity that characterize Chesapeake Bay
make it possible to explore to what extent intrinsic factors are affecting
taphonomic damage, while at the same time controlling for extrinsic factors in the form of habitat type. Although taphonomic bias has been
assessed in a handful of estuarine environments (Jackson, 1968; MacDonald, 1969; Zenetos, 1990, 1991), past studies rarely incorporate information on shell damage. The goals of this study are to: (1) survey the
types of shell damage that occur in the death assemblage of bay benthic
mollusks, and (2) determine whether shell damage in Chesapeake Bay
mollusks varies according to intrinsic factors such as life habit, shell
mineralogy, and shell organic content.
The quality of fossil preservation is controlled by extrinsic factors,
which relate to the environment where specimens accumulate, and intrinsic factors, which relate to the original attributes of the specimens. Extrinsic factors include all the mechanical, biological, and chemical processes acting in a specific environment over a particular timescale. In-
trinsic factors include such biological traits of specimens as life habit,
skeletal mineralogy, skeletal organic content, skeletal thickness, body
plan, and behavior that influence their susceptibility to damage and preservation. During the past two decades, the importance of extrinsic factors
has been emphasized repeatedly (for a review, see Cadee, 1991; Kidwell
and Bosence, 1991; Best and Kidwell, 2000a). Indeed the concept of
taphofacies analysis is predicated on the idea that, as a fossil assemblage
develops, it passes through a taphonomic filter, leaving a distinct signature
that can be used for paleoenvironmental reconstruction (Brett and Baird,
1986; Speyer and Brett, 1986; Davies et al., 1989; Meldahl and Flessa,
1990; Staff and Powell, 1990a; Kowalewski et al., 1994; Pandolfi and
Greenstein, 1997; Best and Kidwell, 2000a). In contrast, the importance
of intrinsic factors in the taphonomic equation (Kidwell and Bosence,
1991) has received relatively little attention until recently (Parsons and
Brett, 1991; Parsons, 1993; Dent, 1995; Aguirre and Farinati, 1999; Best
and Kidwell, 2000b; Kidwell et al., 2001; Nielsen and Funder, 2003;
Lazo, 2004; Nielsen, 2004).
The life habit of an organism is thought to exert some control on shell
damage, depending on exhumation cycles of shells in sediment (Cummins
et al., 1986; Best and Kidwell, 2000b; Kidwell et al., 2001). Intuitively,
epifaunal taxa should be more prone to most forms of damage, since they
are more likely to be deposited postmortem above the sediment-water
interface. Although microbial attack continues after burial, most types of
damage occur exclusively or most rapidly above the sediment-water interface (reviewed by Kidwell and Bosence, 1991; Parsons and Brett,
1991).
Preservation potential of mollusks is also dependent on a variety of
factors related to shell composition, including shell mineralogy, intercrystalline organic content, and crystallite surface area to volume ratio
(Currey and Taylor, 1974; Glover and Kidwell, 1993; Harper, 1998a;
Clark, 1999; Harper, 2000; Kidwell, 2005). In general, aragonitic taxa are
thought to be less likely to be preserved in the fossil record than bimineralic or calcitic ones (Harper, 1998b). Organic content likewise plays
a role in the long-term stability of shell material (Glover and Kidwell,
1993). In the early stages of dissolution, the presence of organic matrix
may temporarily protect mineral crystallites from damage, but over time
organic matrix provides a substrate for microbial decay, ultimately promoting destruction of microstructure (Glover and Kidwell, 1993).
MATERIALS AND METHODS
* Corresponding author.
1 www.sepm.org/archive/index.html
A shell-damage survey was performed on four sites in the main channel of the upper bay (Fig. 1). The sites are distributed lengthwise in the
upper bay and represent two distinct habitat types (Table 1). Sites 1 and
2, located in the northern portion of the upper bay, are characterized by
seasonally variable salinities (2.7 to 15.2 ppt), high silt-clay ratios (92–
95%), and moderate water depth (4.4–7.5 m). In comparison, the southernmost sites (3 and 4) record less variable salinity, sandier substrates,
and shallower water depths (Table 1). The basic differences between these
sites make it possible to control (very generally) for habitat type in these
shell-damage analyses. Death-assemblage material was generously pro-
Copyright 䊚 2006, SEPM (Society for Sedimentary Geology)
0883-1351/06/0021-0442/$3.00
PALAIOS
443
TAPHONOMIC BIAS AND SHELL DAMAGE IN CHESAPEAKE BAY MOLLUSKS
TABLE 2—Damage variables and scoring systems used to assess shell damage in the
death assemblage. Disarticulation was quantified only in taxa with more than one valve
(i.e., bivalves). Periostracum loss was only scored for species with periostraca and was
quantified as follows: 0 signifies that the shell has no periostracum loss (⬎80% of
periostracum intact), 1 signifies that the shell has slight periostracum loss (50–80% of
periostracum intact), 2 signifies that the shell has moderate periostracum loss (10–50%
of periostracum intact), and 3 signifies that the shell has severe periostracum loss
(⬍10% of periostracum intact). Internal and external FSA were scored using a similar
scale: 0 signifies that the shell has no FSA, 1 signifies slight damage (a chalky or dull
appearance), 2 signifies moderate damage (a chalky appearance and ⬍60% of the shell
pitted or eroded), and 3 signifies severe damage (a chalky appearance and ⬎60% of
the surface area pitted or eroded). Fragmentation was scored based on the size of
fragment: 0 signifies no fragmentation (100% of valve), 1 signifies moderate fragmentation (large fragment, ⬎20% of valve), and 2 signifies severe fragmentation (small
fragment, ⬍20% of the valve). Edge modification was ranked similarly: 0 signifies no
edge modification, 1 signifies slight damage (chipped shell edge), and 2 signifies severe
damage (rounded shell edge). Damage variable abbreviations: Enc ⫽ encrustation;
Disart ⫽ disarticulation; Perio ⫽ periostracum loss; Bioero ⫽ bioerosion; FSA ⫽
fine-scale surface alteration; Frag ⫽ fragmentation; Edge mod ⫽ edge modification;
Ext ⫽ external; Int ⫽ internal.
Variable
FIGURE 1—Map of the Chesapeake Bay (Atlantic coast, North America) illustrating the sites sampled in this study.
vided by the benthic monitoring division of the Chesapeake Bay Program
(CBP; Diaz et al., 2003).
The CBP collected one to two samples at each site, in May or September of 2002, using Wildco and hand-operated boxcorers of similar size
(VERSAR, 2002). Field samples were sieved to 0.5 mm and then preserved in a 10% buffered formaldehyde solution. Physical variables, including salinity, temperature, pH, dissolved-oxygen concentration, and
sediment grain size, were collected for each sampling event. Twenty years
worth of sedimentological and physical data were provided for each site
by the CBP (VERSAR, 2002). In the lab, samples were gently resieved
into four fractions (⬎4 mm, 4–2 mm, 2–1 mm, and 1–0.5 mm), rinsed
in freshwater, and air dried. Molluscan material, including both whole
shells and fragments, was picked and identified to species level whenever
possible.
Identifiable shells and fragments were assigned damage states using a
categorical scoring system developed and adapted from previous work
(Davies et al., 1989; Kowalewski et al., 1995; Best and Kidwell, 2000a,
2000b; Kidwell et al., 2001). Data on seven variables summarizing taphonomic shell damage were collected using 10–40x stereoscopic magnification of internal and external shell surfaces (Table 2). Type of encruster
(bryozoan, barnacle, worm, or oyster) or bioeroder (sponge, worm, or
both) was noted. Disarticulation was assessed only in taxa with more than
Position
States
Enc
Disart
Int/Ext
—
Perio
Ext
Bioero
Int/Ext
FSA
Int/Ext
Frag
—
Edge mod
—
% surface
0 ⫽ absent
1 ⫽ present
0 ⫽ none
1 ⫽ slight
2 ⫽ moderate
3 ⫽ severe
0 ⫽ absent
1 ⫽ present
0 ⫽ none
1 ⫽ slight
2 ⫽ moderate
3 ⫽ severe
0 ⫽ none
1 ⫽ ⬍20% shell
2 ⫽ ⬎20% shell
0 ⫽ none
1 ⫽ chipped
2 ⫽ rounded
one valve (i.e., bivalves), and periostracum loss was assessed only for
species with periostraca. SEM (scanning electron microscope) inspection
of shell surface features, which is generally required to distinguish chemical dissolution of crystallites, organic leaching, fungal or algal microboring, and physical abrasion (Cutler, 1995), was not feasible for this
study. Therefore, these factors were combined into a single variable of
fine-scale surface alteration (following Best and Kidwell, 2000a). Shelldamage states with internal scoring were assessed outside of the pallial
line on the interior of the shell and only for bivalve specimens with
identifiable pallial lines. Levels of fragmentation and disarticulation
in this dataset may be artificially inflated due to detritus handling and
TABLE 1—Data on the sites sampled in this study. Data include number of samples collected, sample dates, number of specimens in the death assemblage, salinity, latitude
(lat) and longitude (long), the percentage of silt-clay, the collecting gear, and the water depth (VERSAR, 2002). Note that Sites 1 and 2 (northern sites) are slightly deeper and
have higher silt-clay percentages than Sites 3 and 4 (southern sites).
Site
# Samples
Sample dates
# Specimens
Salinity (ppt)
Lat Long
Silt-clay %
1
2
2
2
3
1
May 02
Sept 02
May 02
Sept 02
Sept 02
677
660
51
184
525
2.7
11.3
9.9
15.2
14.5
94.5
94.4
95.2
92.5
1.9
4
2
May 02
Sept 02
445
670
16.5
14.9
39⬚16.28⬘N
76⬚17.42⬘W
39⬚07.32⬘N
76⬚21.34⬘W
38⬚42.90⬘N
76⬚30.84⬘W
38⬚25.19⬘N
76⬚25.02⬘W
1.0
1.2
Gear
WildCo
Box
WildCo
Box
Modified
Box
Modified
Box
Water depth
(m)
4.4
4.5
7.5
7.5
2.8
2.6
3
444
LOCKWOOD AND WORK
transport (Flessa et al., 1992); however, inspection of samples in the field
and throughout lab preparation indicated that artifactual breakage was not
substantial and should not affect the results significantly (see also
Davies et al., 1989; Best and Kidwell, 2000a). Preservation of the shell
material in formaldehyde by the CBP made it impossible to date the dead
specimens via radiocarbon dating or amino acid racemization. It is unclear, therefore, to what extent shell age is influencing shell damage in
these data.
The size of shells and shell fragments can exert significant influence
over taphonomic signal (Kidwell et al., 2001). The use of multiple sieve
sizes in this study makes it possible to assess the extent to which specimen size is affecting the results. Sample size does differ according to
sieve size (n ⫽ 542 [⬎4 mm]; 762 [4–2 mm]; 1318 [2–1 mm]; 590 [1–
0.5 mm]), suggesting that smaller sieve sizes will drive the results reported here. To explore the effects of sieve size on damage profiles, all
analyses were performed: (1) across all sieve sizes and (2) across larger
sieve sizes (⬎2 mm). The results of these two analyses were similar,
except where noted. A table of species and the sieve fractions in which
they occur is provided in the Supplementary Data1.
Data on life habit were compiled from the literature, and specimens
were divided into epifaunal versus infaunal categories (following Best
and Kidwell, 2000b). Information on shell mineralogy and microstructure
was assembled from a variety of sources (Taylor et al., 1969, 1973; Kidwell and Brenchley, 1996). Data on shell composition were readily available for most of the species represented in the upper bay dataset. In cases
for which compositional data were lacking, data on specimens within the
same genus or family were substituted, but only when data were consistent within that genus or family. Species were divided into two categories
on the basis of shell mineralogy: (1) calcitic, which includes taxa with
purely calcitic and bimineralic shells, and (2) noncalcitic, which includes
taxa with purely aragonitic shells. Species were also divided into two
categories on the basis of shell organic content: (1) low organic content,
which includes taxa with cross-lamellar, complex cross-lamellar, homogeneous aragonite, and foliated calcite microstructures, and (2) high organic content, which includes taxa with nacreous, prismatic aragonite,
and prismatic calcite microstructures (following Glover and Kidwell,
1993).
Whenever possible, field and lab sampling protocol followed recommendations by Kidwell et al. (2001) and Kowalewski and Hoffmeister
(2003) for standardized procedures. Detailed descriptions of field and lab
methodology are provided in Lockwood and Chastant (2006).
RESULTS
Patterns of Shell Damage
Data for each shell-damage variable were summarized in three different
ways: (1) the percentage of specimens affected, (2) the mean damage
score, and (3) the median damage score (excluding encrustation and bioerosion) across all sites. Parsons-Hubbard (2005) has argued that calculating the median for damage scores is more meaningful than calculating
the arithmetic mean, given the discontinuous nature of damage scores. In
the Chesapeake Bay, however, damage due to encrustation and bioerosion
is so rare that calculating the median produces a zero result across all
categories, including species, life habit, mineralogical, and shell organic
content categories. Median damage scores were therefore calculated only
for disarticulation, periostracum loss, FSA, fragmentation, and edge modification. It should also be noted that, for the two-state variables employed
in this study (e.g., disarticulation and bioerosion), percentage and mean
damage metrics yield identical results.
Percent, mean, and median damage metrics are statistically significantly correlated with each other across all damage variables. When summarized at the species level, analyzed via Spearman Rank, and adjusted
for multiple comparisons, R13–17 ranges from 0.56 to 1.0; p ranges from
0.025 to 0.0000001 except for fragmentation mean vs. median score (R17
⫽ 0.45, p ⫽ 0.07). The three metrics also yield very similar results for
PALAIOS
FIGURE 2—Data for each shell damage variable, compiled across all samples and
sites. The data are summarized using three metrics: (1) percentage of specimens
affected (0–100%), (2) mean damage score, and (3) median damage score. Mean
and median damage scores are standardized from 0 to 1 to account for variables
with varying numbers of damage states. Abbreviations for damage scores as in
Table 2.
life-habit and shell-composition analyses; therefore, subsequent discussion will be limited to the results for mean damage scores.
Normality tests (Kolmogorov-Smirnov, Shapiro-Wilk’s W) revealed
that mean damage scores are not normally distributed (d ⫽ ⬍ 0.53, p
⬍ 0.01; W ⫽ 0.83, p ⬍ 0.00001), hence nonparametric statistics were
used for subsequent analyses. Sequential Bonferroni adjustment was applied to each independent pool of statistical analyses (Rice, 1989).
The most common forms of damage (i.e., highest percentage of specimens affected, compiled across all samples and sites) recorded are disarticulation, external FSA, periostracum loss, edge modification, internal
FSA, and fragmentation (Fig. 2). Encrustation and bioerosion are extremely rare in these samples (affecting ⬍4% of specimens; Fig. 2). The
most common encrusters are bryozoans, which cover 0–90% of specimen
surface area, followed by worms (0–60%) and barnacles (0–35%). The
most severe forms of damage (i.e., the highest mean and median) documented are disarticulation, periostracum loss, external FSA, edge modification, fragmentation, and internal FSA (Fig. 2). When these analyses
are restricted to larger sieve sizes (⬎2 mm), the order of the most common and most severe forms of damage remains the same but the order
of the rare forms (encrustation and bioerosion) is altered somewhat. Larger sieve sizes yield statistically significantly greater levels of external
encrustation than smaller (⬍2 mm) sieve sizes (Mann-Whitney U test;
Z1896, 1283 ⫽ ⫺3.54; p ⫽ 0.001), but differences in bioerosion or internal
encrustation are not statistically significant.
Scoring of internal versus external damage in three of the variables
(encrustation, bioerosion, and FSA) makes it possible to compare the two
approaches to assessing shell damage. All three external measures yield
higher levels of damage (1.2 to 8.5x) than internal measures. Both encrustation (R1654 ⫽ 0.65, p ⫽ 0.000001) and FSA (R1650 ⫽ 0.41, p ⫽
0.000001) record statistically significant correlations between internal and
external measures of damage. Bioerosion does not (R1655 ⫽ ⫺0.002, p
⬎ 0.999), although this is due primarily to the dearth of internally bioeroded specimens obtained in these samples (n ⫽ 2). These results did not
change when analyses were limited to larger sieve fractions (⬎2 mm).
Effects of Intrinsic Factors: Data Pooled Across All Sites
Life Habit.—When specimens were pooled across all sites (overall
analysis), Mann-Whitney U tests recorded significant differences between
infaunal and epifaunal specimens for encrustation (both external and internal), periostracum loss, FSA (both external and internal), and fragmentation, after Bonferroni correction (Table 3). Epifaunal specimens display significantly higher levels of damage for external encrustation, internal encrustation, and fragmentation, while infaunal specimens display
significantly higher levels of damage for periostracum loss, external FSA,
PALAIOS
445
TAPHONOMIC BIAS AND SHELL DAMAGE IN CHESAPEAKE BAY MOLLUSKS
TABLE 3—Differences in mean damage score between infaunal and epifaunal specimens for the overall analysis and for each habitat type analysis. Differences are assessed
using Mann-Whitney U tests and results (after Bonferroni adjustment) are provided. Abbreviations for damage scores as in Table 2. Positive Z values denote greater damage in
infaunal relative to epifaunal specimens. Negative Z values denote greater damage in epifaunal relative to infaunal specimens. The number of asterisks denotes the p value for
each test (* ⫽ p ⱕ 0.05; ** ⫽ p ⱕ 0.01; *** ⫽ p ⱕ 0.001; **** ⫽ p ⱕ 0.0001; ***** ⫽ p ⱕ 0.00001). NS (not significant) is used to denote results that are not statistically
significant. NA (not applicable) is used to denote comparisons that could not be assessed due to limited sample size.
Variable
Ext enc
Int enc
Disart
Perio
Ext bioero
Int bioero
Ext FSA
Int FSA
Frag
Edge mod
Overall
Z2804, 408 ⫽ ⫺7.18
****
Z1547, 107 ⫽ ⫺6.59
****
Z2804, 193 ⫽ 1.42
NS
Z2792, 348 ⫽ 4.87
****
Z2798, 408 ⫽ 0.02
NS
Z1550, 105 ⫽ 0.37
NS
Z2798, 408 ⫽ 4.41
****
Z1551, 99 ⫽ 8.53
****
Z2798, 408 ⫽ ⫺5.84
****
Z2798, 382 ⫽ ⫺1.09
NS
and internal FSA (Table 3, Fig. 3). The results for internal damage (i.e.,
encrustation, bioerosion, and FSA) mirror those for external damage (Table 3). Restricting these analyses to specimens from larger sieve fractions
does not affect the results.
Shell Mineralogy.—Mann-Whitney U tests applied to the pooled data
(overall analysis) record significant differences between calcitic and noncalcitic specimens for encrustation (both external and internal), periostracum loss, external FSA, fragmentation, and edge modification, after
Bonferroni correction (Table 4). Calcitic specimens display significantly
higher levels of damage for external encrustation, internal encrustation,
fragmentation, and edge modification, while noncalcitic specimens display significantly higher levels of damage for periostracum loss and external FSA (Table 4, Fig. 4). When these analyses were limited to large
sieve size data, the results remained the same.
Shell Organic Content.—Combining data from all specimens across all
sites (overall analysis) yielded significant differences between specimens
FIGURE 3—Bar graph illustrating the mean (⫾SE) damage score for epifaunal
versus infaunal specimens (compiled across all samples and sites, overall analysis)
for a subset of the damage variables scored. Abbreviations for damage scores as in
Table 2. In the overall analysis, epifaunal specimens have significantly higher damage scores for fragmentation, while infaunal specimens have significantly higher
damage scores for periostracum loss, external FSA, and internal FSA. Asterisks
denote statistically significant differences.
Southern sites
Z1407, 233 ⫽ ⫺8.18
*****
Z890, 41 ⫽ ⫺8.08
*****
Z1407, 183 ⫽ 1.66
NS
Z1395, 173 ⫽ 17.71
*****
Z1401, 233 ⫽ ⫺0.01
NS
Z892, 39 ⫽ 0.3
NS
Z1401, 233 ⫽ 7.2
*****
Z891, 40 ⫽ 2.47
**
Z1401, 233 ⫽ ⫺11.04
*****
Z1401, 233 ⫽ ⫺3.17
*
Northern sites
Z1397, 175 ⫽ ⫺0.43
NS
NA
Z1397, 10 ⫽ 0.24
NS
Z1397, 175 ⫽ ⫺9.61
*****
Z1397, 175 ⫽ 0.35
NS
NA
Z1397, 175 ⫽ ⫺0.52
NS
Z660, 59 ⫽ 9.12
*****
Z1397, 175 ⫽ 2.41
NS
Z1397, 149 ⫽ 2.44
NS
with low versus high shell organic content for external and internal encrustation, disarticulation, external FSA, fragmentation, and edge modification, after Bonferroni correction of Mann-Whitney U tests (Table 5).
Specimens with high shell organic content display significantly higher
levels of damage for external and internal encrustation, disarticulation,
fragmentation, and edge modification, while specimens with low shell
organic content display significantly higher levels of damage for external
FSA (Table 5, Fig. 5). Restricting these analyses to larger sieve fractions
resulted in statistically significantly higher levels of periostracum loss in
specimens with high shell organic content (Mann-Whitney U test,
Z1079, 177 ⫽ ⫺2.88; p ⫽ 0.04).
Effects of Intrinsic Factors: Controlling for Habitat Type
To control for extrinsic factors, the above analyses were performed
separately on each of the two habitat types (southern versus northern)
sampled in this study. As described above, Sites 3 and 4 (southern) represent a slightly shallower habitat with a coarser grained substrate and
less variable salinity than Sites 1 and 2 (northern). For the sake of simplicity, these two habitat types are referred to as northern versus southern
as opposed to deeper- and finer-grained sediment versus shallower- and
coarser-grained sediment. Sampling of the southern sites produced more
material than sampling of the northern sites; therefore, the results for the
overall analysis consistently match those of the southern site analysis
(Tables 3, 4, and 5).
Life Habit.—The only difference between the results obtained for the
overall analysis and those obtained for the southern sites involves edge
modification, which is significantly more severe in epifaunal relative to
infaunal specimens at the latter. The northern sites recorded substantially
fewer differences between infaunal and epifaunal levels of damage (Table
3). While the difference for internal FSA remained significant, the differences for external encrustation, external FSA, and fragmentation did
not. Although both habitat types document a statistically significant difference between epifaunal and infaunal levels of periostracum loss, infaunal damage is greater in the southern sites and lower in the northern
sites. The absence of internally encrusted or bioeroded specimens in the
northern sites make it impossible to test for differences in these two
damage variables while controlling for habitat type. The only statistically
significant difference between infaunal and epifaunal damage that is
446
PALAIOS
LOCKWOOD AND WORK
TABLE 4—Differences in mean damage score between calcitic and non-calcitic specimens for the overall analysis and for each habitat type analysis. Differences are assessed
using Mann-Whitney U tests and results (after Bonferroni adjustment) are provided. Damage score abbreviations follow Table 2. Positive Z values denote greater damage in
noncalcitic relative to calcitic specimens. Negative Z values denote greater damage in calcitic relative to noncalcitic specimens. The number of asterisks denotes the p value for
each test (* ⫽ p ⱕ 0.05; ** ⫽ p ⱕ 0.01; *** ⫽ p ⱕ 0.001; **** ⫽ p ⱕ 0.0001; ***** ⫽ p ⱕ 0.00001). NS (not significant) is used to denote results that are not statistically
significant. NA (not applicable) is used to denote comparisons that could not be assessed due to limited sample size.
Variable
Ext enc
Int enc
Disart
Perio
Ext bioero
Int bioero
Ext FSA
Int FSA
Frag
Edge mod
Overall
Z3019, 190 ⫽ ⫺11.32
****
Z1617, 36 ⫽ ⫺11.62
****
Z2804, 190 ⫽ 1.41
NS
Z2964, 173 ⫽ 12.31
****
Z3013, 190 ⫽ ⫺0.79
NS
Z1620, 34 ⫽ 0.20
NS
Z3013, 190 ⫽ 5.89
****
Z1614, 35 ⫽ 1.63
NS
Z3013, 190 ⫽ ⫺12.63
****
Z2987, 190 ⫽ ⫺7.05
****
recorded consistently across both habitat types is the higher level of internal FSA documented in infaunal versus epifaunal specimens.
Shell Mineralogy.—As expected, the southern sites generated results
similar to the overall analysis (Table 4), the only difference being
significantly greater internal FSA in noncalcitic specimens in the southern
sites. The northern sites recorded weak, generally nonsignificant differences between noncalcitic and calcitic specimens (Table 4), with the exception of external encrustation. Calcitic specimens consistently exhibit
significantly higher levels of external encrustation than noncalcitic ones
across both habitat types.
Shell Organic Content.—On the whole, the southern sites produced
patterns similar to the overall analysis; the patterns were in the same
direction although the statistical significance varied. In the southern sites,
differences in disarticulation are not statistically significant, and specimens with low shell organic display significantly higher levels of perios-
Southern sites
Northern sites
Z1457, 183 ⫽ ⫺9.39
*****
Z900, 31 ⫽ ⫺9.34
*****
Z1407, 183 ⫽ 1.66
NS
Z1402, 166 ⫽ 18.3
*****
Z1451, 183 ⫽ ⫺0.26
NS
Z902, 29 ⫽ 0.25
NS
Z1451, 183 ⫽ 8.51
*****
Z901, 30 ⫽ 3.46
**
Z1451, 183 ⫽ ⫺13.63
*****
Z1451, 183 ⫽ ⫺6.02
*****
Z1562, 7 ⫽ ⫺5.98
*****
NA
Z1397, 7 ⫽ 0.21
NS
Z1562, 7 ⫽ 1.52
NS
Z1562, 7 ⫽ 0.07
NS
NA
Z1562, 7 ⫽ 2.55
NS
Z713, 5 ⫽ 0.29
NS
Z1562, 7 ⫽ ⫺2.13
NS
Z1536, 7 ⫽ ⫺1.70
NS
tracum loss than specimens with high shell organic content (Table 5). The
differential disarticulation of specimens with high shell organic content
recorded in the overall analysis is driven by specimens from the northern
sites. Interestingly, the patterns for periostracum loss and external FSA
in the southern sites, in which low show significantly greater damage
than high, are completely opposite those recorded in the northern sites.
The only significant differences documented consistently across both habitat types are fragmentation and edge modification, which are significantly
more severe for specimens with high shell organic content.
Shell-damage patterns for specimens with noncalcitic shell mineralogy
(Fig. 4), those with low shell organic content (Fig. 5), and those with
infaunal life habits (Fig. 3) mimic each other primarily because the vast
majority of infaunal taxa in the Chesapeake Bay are characterized by
noncalcitic shell mineralogy (100%) and low shell organic content (75%).
DISCUSSION
Shell Damage
FIGURE 4—Bar graph illustrating the mean (⫾SE) damage score for specimens in
different shell mineralogical categories (compiled across all samples and sites, overall analysis) for a subset of the damage variables scored. Abbreviations for damage
scores as in Table 2. Asterisks denote statistically significant differences. In the
overall analysis, calcitic specimens show significantly more severe damage due to
fragmentation and edge modification than noncalcitic specimens, while the reverse
is true for periostracum loss and external FSA.
This survey of molluscan shell damage in the upper Chesapeake Bay
identified disarticulation, external FSA, periostracum loss, edge modification, internal FSA, and fragmentation as the most common and severe
forms of damage. These results agree well with Kidwell and Bosence’s
(1991) review of shell damage and its distribution across estuarine environments. They noted that, in these settings, dissolution is common,
while bioerosion, abrasion, and fragmentation are often present. Kidwell
and Bosence (1991) mentioned that specimens can also be found articulated and in situ in these environments, although it is rare in the upper
bay.
The salinity fluctuations, shape, and latitude of the Chesapeake Bay
make it difficult to identify an appropriate analogue for comparison of
shell-damage profiles; few studies have examined taphonomic signatures
in estuarine or marginal marine environments. For those studies that have,
damage levels are generally comparable to those levels documented in
this study. For example, the disarticulation levels documented in the
Chesapeake Bay are extremely high (99%), but they correspond closely
with the results obtained for Henderson and Frey (1986) for the northern
flank of Doboy Sound (85–100%) near Sapelo Island, Georgia. FSA is
considerably more common in the upper bay (external 94%, internal 79%)
PALAIOS
447
TAPHONOMIC BIAS AND SHELL DAMAGE IN CHESAPEAKE BAY MOLLUSKS
TABLE 5—Differences in mean damage score between low and high shell organic content specimens for the overall analysis and for each habitat type analysis. Differences are
assessed using Mann-Whitney U tests, and results (after Bonferroni adjustment) are provided. Damage score abbreviations follow Table 2. Positive Z values denote greater
damage in low-relative to high-shell-organic-content specimens. Negative Z values denote greater damage in high relative to low shell organic content specimens. The number
of asterisks denotes the p value for each test (* ⫽ p ⱕ 0.05; ** ⫽ p ⱕ 0.01; *** ⫽ p ⱕ 0.001; **** ⫽ p ⱕ 0.0001; ***** ⫽ p ⱕ 0.00001). NS (not significant) is used
to denote results that are not statistically significant. NA (not applicable) is used to denote comparisons that could not be assessed due to limited sample size.
Variable
Ext enc
Int enc
Disart
Perio
Ext bioero
Int bioero
Ext FSA
Int FSA
Frag
Edge mod
Overall
Z2645, 345 ⫽ ⫺7.64
*****
Z1530, 51 ⫺9.49
*****
Z2645, 345 ⫽ ⫺5.05
*****
Z2620, 341 ⫽ 1.92
NS
Z2643, 341 ⫽ ⫺0.10
NS
Z1533, 49 ⫽ 0.25
NS
Z2643, 341 ⫽ 3.04
*
Z1535, 49 ⫽ 1.27
NS
Z2643, 341 ⫽ ⫺16.55
*****
Z2643, 341 ⫽ ⫺6.29
*****
than in Doboy Sound (defined as loss of color, gloss, periostracum, or
ligament in ‘‘old’’ shells, 12%) or in San Luis Pass, a tidal inlet near
Galveston Island (defined as abrasion, 63%, and dissolution, 28%; Davies
et al., 1989). The increased levels of FSA recorded in the bay could be
a result of the slightly acidic nature of the bay’s brackish water, which
would be expected to etch and leach carbonate shells (Alexandersson,
1972; Pip, 1988).
Levels of fragmentation documented in the upper bay (67%) are very
similar to those observed by Davies et al. (1989) in San Luis inlet (68%).
Bay levels of encrustation and bioerosion (external encrustation 0.53%,
internal encrustation 0.2%, external bioerosion 3.4%, internal bioerosion
0.4%) are also a close match with San Luis inlet levels for bivalves
(defined as epibionts, borings, and traces, 2%) but not gastropods (44%;
Davies et al., 1989). The lower level of damage observed for gastropods
in the Chesapeake Bay (effectively 0%) is a pattern that Best and Kidwell
Southern sites
Z1382, 201 ⫽ ⫺8.72
*****
Z891, 26 ⫽ ⫺10.15
*****
Z1382, 201 ⫽ ⫺2.20
NS
Z1357, 197 ⫽ 16.16
*****
Z1380, 197 ⫽ ⫺0.148
NS
Z893, 24 ⫽ 0.23
NS
Z1380, 197 ⫽ 7.51
*****
Z893, 24 ⫽ 2.49
NS
Z1380, 197 ⫽ ⫺14.61
*****
Z1380, 197 ⫽ ⫺5.19
*****
Northern sites
Z1263, 144 ⫽ ⫺0.52
NS
NA
Z1263, 144 ⫽ ⫺6.06
*****
Z1263, 144 ⫽ ⫺12.87
*****
Z1263, 144 ⫽ 0.34
NS
NA
Z1263, 144 ⫽ ⫺3.21
*
Z642, 25 ⫽ ⫺1.98
NS
Z1263, 144 ⫽ ⫺9.13
*****
Z1263, 144 ⫽ ⫺3.51
**
(2000a) attributed in their Panama study to limited exposure of shells
above the sediment-water interface. In the present study, the sieve size
used also plays a role, because smaller specimens and finer sieve sizes
tend to show significantly lower frequencies of boring and encrustation
(Kidwell et al., 2001; Kowalewski and Hoffmeister, 2003; Rodland et al.,
2004).The substantial salinity variation in the bay is also likely to be a
factor, inasmuch as many of the taxa responsible for such damage vary
in abundance in response to salinity gradients (e.g., Buatois et al., 2005).
Internal damage is generally assumed to be postmortem in origin (but
see Lazo, 2004), while epifaunal damage can accrue both pre- and postmortem, leading some past workers to advocate the use of internal damage as an unambiguous measure of postmortem damage in taphofacies
studies (Best and Kidwell, 2000a, 2000b). The present study is one of a
number of studies to assess both external and internal measures of shell
damage (Davies et al., 1989; Staff and Powell, 1990b; Callender and
Powell, 1992; Callender et al., 1992; Kowalewski et al., 1994; Kowalewski et al., 1995; Kidwell et al., 2001). The majority of these studies
have documented equivalent or greater damage on external surfaces for
encrustation (Kidwell et al., 2001), bioerosion (Kidwell et al., 2001), and
FSA (Davies et al., 1989; Staff and Powell, 1990b; Callender et al., 1992;
Kowalewski et al., 1995; Kidwell et al., 2001), similar to the results
reported here. The results for the upper Chesapeake Bay also indicate
that, with the exception of bioerosion, internal scoring was significantly
correlated with external scoring in the upper bay. This suggests that, for
environments similar to those sampled here, external damage may provide a useful proxy for internal damage.
Does Shell Damage Vary According to Life Habit?
FIGURE 5—Bar graph illustrating the mean (⫾SE) damage score for specimens in
different shell organic categories (compiled across all samples and sites, overall
analysis) for a subset of the damage variables scored. Abbreviations for damage
scores as in Table 2. Asterisks denote statistically significant differences. In the
overall analysis, damage due to disarticulation, fragmentation, and edge modification
is significantly more severe for specimens with high shell organic content. Specimens
with low shell organic content record significantly more severe damage due to external FSA.
Comparisons of epifaunal versus infaunal levels of shell damage in the
upper Chesapeake Bay yielded one consistent difference that was statistically significant across both habitat types. Infaunal specimens exhibit
more severe internal FSA than epifaunal ones. If external damage represents both pre- and postmortem processes, while internal damage is
limited to postmortem processes (but see Lazo, 2004), it follows that
epifaunal taxa should show more severe external damage than infaunal.
While this is the case for encrustation, fragmentation, and edge modification in the southern sites in the bay, it does not hold true across all
damage variables or across all habitat types.
448
LOCKWOOD AND WORK
The majority of taphonomic studies that have compared shell damage
in epifaunal versus infaunal specimens have documented higher damage
levels in the former (Callender et al., 1990; Parsons and Brett, 1991;
Parsons, 1993; Dent, 1995; Aguirre and Farinati, 1999; Best and Kidwell,
2000b; Kidwell et al., 2001; Lazo, 2004; Nielsen, 2004). It is worth
noting that many of these studies (e.g., Parsons and Brett, 1991; Parsons,
1993; Dent, 1995) have focused predominantly on hardground as opposed
to the exclusively soft-sediment environments sampled here. When Best
and Kidwell (2000b) compared epifaunal versus infaunal damage in softsubstrate environments in Panama, they recorded no significant difference
between the two. Similarly, the analyses of Kidwell et al. (figure 9 in
2001) of inlet floor muds in Panama showed no significant difference in
shell damage due to bivalve life habit. Figure 9 of Kidwell et al. (2001)
demonstrates that differences between epifaunal and infaunal shell damage vary according to environment and are rarely statistically significant.
Few studies have attempted to control for habitat while assessing the
effect of life habit on shell damage. Aguirre and Farinati (1999) demonstrate that infaunal bivalve species from the late Quaternary of Argentina are better preserved than epifaunal taxa, but they also note that the
former inhabit soft bottom environments whereas the latter inhabit hardbottom environments. Since coarser-grain-size environments are often
characterized by higher levels of shell damage (e.g., Best and Kidwell,
2000a), it is difficult to determine whether the differences in shell damage
documented by Aguirre and Farinati (1999) are due to differences in life
habit or in environment. More recently, Lazo (2004) documented a higher
proportion of discoloration, corrasion (internal and external), internal encrustation, and bioerosion in epifaunal relative to infaunal specimens of
Protothaca staminea from San Juan Island, Washington. This study is
similarly complicated by the fact that infaunal Protothaca are limited to
mud and muddy sand habitats while epifaunal Protothaca are restricted
to gravels.
The majority of past studies have also focused on external as opposed
to internal damage, with the exception of Best and Kidwell (2000b),
Kidwell et al. (2001), and Lazo (2004). In the upper Chesapeake Bay,
infaunal specimens exhibit significantly higher levels of internal FSA than
epifaunal specimens. Best and Kidwell (2000b) also found that, although
the difference was not statistically significant, infaunal taxa tended to
have more severe internal FSA in muddy environments in siliciclastic
regimes in Panama.
Returning to the results for the upper Chesapeake Bay, it is interesting
to compare the direction of differential damage according to life habit.
In the southern sites, four damage variables show greater damage in epifaunal specimens while three show the opposite (Table 3). In the northern
sites, one damage variable records preferential damage in one direction,
while another records it in the opposite direction. These patterns seem to
suggest that taphonomic bias in one direction (e.g., epifaunal more damaged than infaunal specimens), which could lead to the selective removal
of specimens with that particular life habit, would be effectively canceled
out by bias in the opposite direction. This assumes, of course, that all
damage variables are equally likely to lead to taphonomic bias, an assumption that has yet to be tested in the Chesapeake Bay or other estuarine environments.
The results for the Chesapeake Bay, when considered with past studies,
suggest that differential damage according to life habit (either the direction or statistical significance) varies drastically according to the damage
variable and the habitat type examined. Contrary to expectations, epifaunal specimens rarely show higher levels of external or internal damage
than infauna in soft-sediment environments. In the upper bay, differential
damage of a particular life habit is unlikely to lead to a strong taphonomic
bias in either direction.
Does Shell Damage Vary According to Shell Composition?
Comparisons of shell damage in calcitic versus noncalcitic specimens
in the upper Chesapeake Bay yielded a single consistent statistical dif-
PALAIOS
ference, when the results were controlled for habitat type. Calcitic specimens, all of which are epifaunal, experience higher levels of external
encrustation. Statistically significant differences between specimens with
low versus high shell organic content were documented consistently for
two damage variables across both habitat types. Specimens with high
shell organic content record significantly greater damage due to fragmentation and edge modification than specimens with low shell organic
content.
Few studies have explicitly assessed whether shell damage differs according to shell composition. Best and Kidwell (2000b) divided modern
Panamanian bivalves into three categories on the basis of shell composition: (1) calcitic: equivalent to combining the calcitic and low-shellorganic-content categories in this study, (2) porcellaneous aragonitic:
equivalent to combining the noncalcitic and low-shell-organic-content
categories, and (3) high organic: equivalent to the high-shell-organiccontent category. Best and Kidwell (2000b) found no significant differences in damage frequencies for calcitic versus porcellaneous aragonitic
shells in soft-sediment environments. When the calcitic and low-shellorganic-content specimens from the upper Chesapeake Bay were compared to the noncalcitic and high-shell-organic-content specimens in the
current study, Mann-Whitney U tests revealed only one significant difference after Bonferroni adjustment. Calcitic low-shell-organic-content
specimens exhibited significantly more damage due to edge modification
than noncalcitic low-shell-organic-content specimens (Z2626, 17 ⫽ ⫺4.61,
p ⫽ 0.004 after Bonferroni adjustment).
The dearth of a strong and consistent link between shell damage and
mineralogy across both habitat types sampled in this study is perhaps
not surprising. A general link between shell mineralogy and preservation in mollusks has been postulated for decades, but it is surprisingly
difficult to document in the fossil record or via experimental approaches. One commonly quoted statistic contends that 94% of partially and
fully calcitic living molluscan genera are recorded from the fossil record, compared with only 82% of fully aragonitic living genera (Harper,
1998b). Recent reanalysis of these same data at the family level suggests that entirely aragonitic heteroconch bivalves actually have a more
complete record than calcite-bearing pteriomorphs (Harper, 1998b; see
supplementary online material in Kidwell, 2005). An analysis of shellbed thickness throughout the Phanerozoic revealed that shell mineralogy appears to exert little influence over shell-bed thickness (Kidwell
and Brenchley, 1996). Similarly, a comparison of genus duration in
fossil marine bivalves and shell composition, including mineralogy and
shell organic content, shows little relationship between the two (Kidwell, 2005).
Experimental approaches have suggested that aragonite has a lower
thermodynamic stability than calcite in seawater (Morse et al., 1980;
Mucci, 1983; Sanders, 2003). Unfortunately, even this basic observation
breaks down with variation in ambient seawater chemistry and microstructural surface area (Flessa and Brown, 1983; Walter and Morse, 1984;
Walter, 1985; Walter and Morse, 1985). Crystallite size, magnesium content of calcites, microbial activity, periostracal coverings, and organic
matrix also play complex roles in the determination of shell dissolution
rates (Walter and Morse, 1984; Heinrich and Wefer, 1986; Morse, 1986;
Bischoff et al., 1987; Hunter, 1990; Simon and Poulicek, 1990; Glover
and Kidwell, 1993; Harper, 2000).
The higher levels of fragmentation and edge modification documented
in high-shell-organic-content specimens in the upper Chesapeake Bay do
complement the results that Zuschin and Stanton (2001) obtained in their
experiments linking shell weight to strength. They found that, as organicrich shells degrade through time, they become progressively weaker. Similarly, Lescinsky et al. (2002) noted high frequencies of edge chipping in
specimens with high shell organic content. The only differences between
specimens with low and high shell organic content recorded by Best and
Kidwell (2000b) in soft-sediment environments in Panama were significantly higher levels of edge rounding and internal surface degradation in
high-organic-content shells in sandy mud environments.
PALAIOS
TAPHONOMIC BIAS AND SHELL DAMAGE IN CHESAPEAKE BAY MOLLUSKS
It is difficult to imagine how the differential damage according to shell
mineralogy documented in the upper Chesapeake Bay could lead to the
selective removal of specimens with either calcitic or noncalcitic mineralogy. In the southern sites, four variables show higher levels of damage
in calcitic relative to noncalcitic specimens, while three variables show
the opposite pattern (Table 4). In the northern sites, only one variable,
external encrustation, shows higher levels of damage in calcitic specimens. Once again, any taphonomic bias established in one direction (e.g.,
calcitic specimens more damaged than noncalcitic) is likely to be negated
by bias in the other direction. This is not necessarily the case for shell
organic content (Table 5). In the northern sites, five damage variables
record greater damage in high-, as opposed to low-, shell-organic-content
specimens. In the southern sites, four variables show greater damage in
specimens with high versus low shell organic content, while two show
the opposite. In this analysis, damage of high-shell-organic-content specimens far outweighs damage of low-shell-organic-content specimens, at
least in northern sites. This suggests that a taphonomic filter, based on
shell organic content, may exist in the upper Chesapeake Bay. It would
be interesting to explore to what extent different types and levels of
damage lead to taphonomic bias in these soft-sediment estuarine environments.
All in all, the results documented here and summarized from past work
indicate that differential damage (the direction or statistical significance)
according to shell mineralogy can vary greatly according habitat type and
the damage variable scored. There is little evidence to support the preferential damage of or taphonomic bias against aragonitic specimens, at
least in soft-sediment environments. The results for shell organic content
are equally variable, although some evidence does suggest that specimens
with high shell organic content are more prone to fragmentation and
breakage. In the northern sites of the upper Chesapeake Bay, the greater
levels of damage documented in high-shell-organic-content specimens
may lead to selective removal of these taxa from these sites.
CONCLUSIONS
A survey of signatures of taphonomic damage in the upper Chesapeake
Bay revealed that the most severe forms of shell damage are disarticulation, periostracum loss, external FSA, edge modification, fragmentation,
and internal FSA. Encrustation and bioerosion are exceedingly rare in
these assemblages. Internal and external surface measures of damage are
statistically significantly correlated for encrustation and FSA, suggesting
that, in these particular environments, external measures may provide a
useful proxy for internal ones.
Shell damage differs according to both life habit and shell composition,
but the significance and direction of differential damage tends to vary
according to the type of damage assessed or the habitat type sampled.
For example, periostracum loss is significantly greater in infaunal specimens and specimens with low shell organic content in the southern sites
and significantly lower in the northern sites. Similarly, external FSA is
significantly lower in specimens with low shell-organic content in the
northern sites and significantly greater in the southern sites. The only
patterns of differential damage that consistently were statistically significant across both habitat types were internal FSA, external encrustation,
and fragmentation and edge modification. Contrary to expectations, infaunal specimens exhibit significantly higher levels of internal FSA than
epifaunal specimens. Along the same lines, calcitic specimens show greater external encrustation than noncalcitic specimens. High-shell-organiccontent specimens exhibit significantly greater fragmentation and edge
rounding than low-shell-organic-content specimens, regardless of habitat
type. The only pattern of differential damage that might lead to a taphonomic filter in the upper Chesapeake Bay is the preferential damage of
high-shell-organic-content specimens in the northern sites.
ACKNOWLEDGMENTS
This study would not have been possible without the death-assemblage
material generously provided by D. Dauer (CBP), B. Rodi (CBP), R.
449
Llanso (CBP), and R. Seitz (VIMS). We also thank S. Kidwell for providing shell mineralogy data. T. Rothfus and S. Kidwell helped design
the study. J. Curry, M. Oreska, K. Leary, and C. Compton provided lab
assistance. M. Kowalewski, K. Parsons-Hubbard, T. Rothfus, L. Chastant,
E. Gercke, and J. Swaddle commented on the manuscript. K. Farnsworth
and S. Harris designed Figure 1. We thank the Jeffress Memorial Trust
and W&M for financial support.
REFERENCES
AGUIRRE, M.L., and FARINATI, E.A., 1999, Taphonomic processes affecting late Quaternary molluscs along the coastal area of Buenos Aires Province: Palaeogeography, Palaeoclimatology, Palaeoecology, v. 149, p. 283–304.
ALEXANDERSSON, T., 1972, Intragranular growth of marine aragonite and Mg-calcite:
Evidence of precipitation from supersaturated seawater: Journal of Sedimentary
Petrology, v. 42, p. 441–460.
BEST, M.M.R., and KIDWELL, S.M., 2000a, Bivalve taphonomy in tropical mixed
siliciclastic-carbonate settings; I, Environmental variation in shell condition: Paleobiology, v. 26, p. 80–102.
BEST, M.M.R., and KIDWELL, S.M., 2000b, Bivalve taphonomy in tropical mixed
siliciclastic-carbonate settings; II, Effect of bivalve life habits and shell types:
Paleobiology, v. 26, p. 103–115.
BISCHOFF, W.D., MACKENZIE, F.T., and BISHOP, F.C., 1987, Stabilities of synthetic magnesium calcites in aqueous solution: Comparison with biogenic materials: Geochimica et Cosmochimica Acta, v. 51, p. 1,413–1,423.
BRETT, C.E., and BAIRD, G.C., 1986, Comparative taphonomy: A key to paleoenvironment interpretation based on fossil preservation: PALAIOS, v. 1, p. 207–227.
BUATOIS, L.A., GINGRAS, M.K., MACEACHERN, J., MÁNGANO, M.G., ZONNEVELD, J., PEMBERTON, S.G., NETTO, R.G., and MARTIN, A., 2005, Colonization of brackish-water
systems through time: Evidence from the trace-fossil record: PALAIOS, v. 20, p.
321–347.
CADEE, G.C., 1991, The history of taphonomy, in Donovan, S.K., ed., The Processes
of Fossilization: New York, Columbia University Press, p. 3–21.
CALLENDER, W.R., and POWELL, E.N., 1992, Taphonomic signature of petroleum seep
assemblages on the Louisiana upper continental slope: Recognition of autochthonous shell beds in the fossil record: PALAIOS, v. 7, p. 388–408.
CALLENDER, W.R., POWELL, E.N., STAFF, G.M., and DAVIES, D.J., 1992, Distinguishing
autochthony, parautochthony and allochthony using taphofacies analysis: Can cold
seep assemblages be discriminated from assemblages of the nearshore and continental shelf?: PALAIOS, v. 7, p. 409–421.
CALLENDER, W.R., STAFF, G.M., POWELL, E.N., and MACDONALD, I.R., 1990, Gulf of
Mexico hydrocarbon seep communities. V. Biofacies and shell orientation of autochthonous shell beds below storm wave base: PALAIOS, v. 5, p. 2–14.
CLARK, G.R., 1999, Organic matrix taphonomy in some molluscan shell microstructures: Palaeogeography, Palaeoclimatology, Palaeoecology, v. 149, p. 305–312.
CUMMINS, H., POWELL, E.N., STANTON, R.J. JR, and STAFF, G., 1986, The size frequency distribution in palaeoecology: Effects of taphonomic processes during formation of molluscan death assemblages in Texas Bays: Palaeontology, v. 29, p.
495–518.
CURREY, J.D., and TAYLOR, J.D., 1974, The mechanical behaviour of some molluscan
hard tissues: Journal of Zoology, v. 173, p. 395–406.
CUTLER, A.H., 1995, Taphonomic implications of shell surface textures in Bahı́a la
Choya, northern Gulf of California: Palaeogeography, Palaeoclimatology, Palaeoecology, v. 114, p. 219–240.
DAVIES, D.J., POWELL, E.N., and STANTON, J.R.J., 1989, Taphonomic signature as a
function of environmental process: Shells and shell beds in a hurricane-influenced
inlet on the Texas coast: Palaeogeography, Palaeoclimatology, Palaeoecology, v.
72, p. 317–356.
DENT, S.R., 1995, A taphofacies model of the recent South Florida continental shelf:
A new perspective for a classic, exposed carbonate environment: Unpublished
Ph.D. thesis, University of Cincinnati, Cincinnati, 242 p.
DIAZ, R.J., CUTTER, G.R., and DAUER, D.M., 2003, A comparison of two methods for
estimating the status of benthic habitat quality in the Virginia Chesapeake Bay:
Journal of Experimental Marine Biology and Ecology, v. 285–286, p. 371–381.
FLESSA, K., and BROWN, T.J., 1983, Selective solution of macroinvertebrate calcareous
hard parts: Lethaia, v. 16, p. 193–205.
FLESSA, K., KOWALEWSKI, M., and WALKER, S.E., 1992, Post-collection taphonomy:
Shell destruction and the Chevrolet: PALAIOS, v. 7, p. 553–554.
GLOVER, C.P., and KIDWELL, S.M., 1993, Influence of organic matrix on the postmortem destruction of molluscan shells: Journal of Geology, v. 101, p. 729–747.
HARPER, E.M., 1998a, Calcite in chamid bivalves: Journal of Molluscan Studies, v.
64, p. 391–399.
HARPER, E.M., 1998b, The fossil record of bivalve molluscs, in Donovan, S.K., and
Paul, C.R.C., eds., The Adequacy of the Fossil Record: New York, John Wiley
and Sons, p. 243–267.
450
PALAIOS
LOCKWOOD AND WORK
HARPER, E.M., 2000, Are calcitic layers an effective adaptation against shell dissolution in the Bivalvia?: Journal of Zoology, v. 251, p. 179–186.
HEINRICH, R., and WEFER, G., 1986, Dissolution of biogenic carbonates: Effect of
skeletal structure: Marine Geology, v. 71, p. 341–362.
HENDERSON, S.W., and FREY, R.W., 1986, Taphonomic redistribution of mollusk shells
in a tidal inlet channel, Sapelo Island, Georgia: PALAIOS, v. 1, p. 3–16.
HUNTER, R.D., 1990, Shell dissolution rate in freshwater pulmonate snails: Influence
of periostracum and water chemistry: Nautilus, v. 104, p. 87–91.
JACKSON, J.B.C., 1968, Neontological and paleontological study of the autecology
and synecology of the molluscan fauna of Fleets Bay, Virginia: Unpublished M.S.
thesis, George Washington University, 111 p.
KIDWELL, S.M., 2005, Shell composition has no net impact on large-scale evolutionary patterns in mollusks: Science, v. 307, p. 914–917.
KIDWELL, S.M., and BOSENCE, D.W.J., 1991, Taphonomy and time-averaging of marine shelly faunas, in Allison, P.A., and Briggs, D.E.G., eds., Taphonomy: Releasing the Data Locked in the Fossil Record: New York, Plenum Press, p. 115–
209.
KIDWELL, S.M., and BRENCHLEY, P.J., 1996, Evolution of the fossil record: Thickness
trends in marine skeletal accumulations and their implications, in Jablonski, D.,
Erwin, D.H., and Lipps, J.H., eds., Evolutionary Paleobiology: Essays in Honor
of James W. Valentine: Chicago, University of Chicago Press, p. 290–336.
KIDWELL, S.M., ROTHFUS, T.A., and BEST, M.M.R., 2001, Sensitivity of taphonomic
signatures to sample size, sieve size, damage scoring system, and target taxa:
PALAIOS, v. 16, p. 26–52.
KOWALEWSKI, M., FLESSA, K.W., and AGGEN, J.A., 1994, Taphofacies analysis of recent shelly cheniers (beach ridges), northeastern Baja California, Mexico: Facies,
v. 31, p. 209–241.
KOWALEWSKI, M., FLESSA, K.W., and HALLMAN, D.P., 1995, Ternary taphograms: Triangular diagrams applied to taphonomic analysis: PALAIOS, v. 5, p. 478–483.
KOWALEWSKI, M., and HOFFMEISTER, A.P., 2003, Sieves and fossils: Effects of mesh
size on paleontological patterns: PALAIOS, v. 18, p. 460–469.
LAZO, D.G., 2004, Bivalve taphonomy: Testing the effect of life habits on the shell
condition of the Littleneck clam Protothaca (Protothaca) staminea (Mollusca: Bivalvia): PALAIOS, v. 19, p. 451–459.
LESCINSKY, H.I., EDINGER, E., and RISK, M.J., 2002, Mollusc shell encrustation and
bioerosion rates in a modern epeiric sea: Taphonomy experiments in the Java Sea,
Indonesia: PALAIOS, v. 17, p. 171–191.
LOCKWOOD, R., and CHASTANT, L.R., 2006, Quantifying taphonomic bias of compositional fidelity, species richness, and rank abundance in molluscan death assemblages from the upper Chesapeake Bay: PALAIOS, 21(4): 376–383.
MACDONALD, K.B., 1969, Quantitative studies of salt marsh mollusk faunas from the
North American Pacific Coast: Ecological Monographs, v. 39, p. 33–60.
MELDAHL, K.H., and FLESSA, K.W., 1990, Taphonomic pathways and comparative
biofacies and taphofacies in a recent intertidal/shallow shelf environment: Lethaia,
v. 23, p. 43–60.
MORSE, J.W., 1986, The surface chemistry of calcium carbonate minerals in natural
waters: An overview: Marine Chemistry, v. 20, p. 91–112.
MORSE, J.W., MUCCI, A., and MILLERO, F.J., 1980, The solubility of calcite and aragonite in seawater of 35‰ salinity at 25C and atmospheric pressure: Geochimica
et Cosmochimica Acta, v. 44, p. 85–94.
MUCCI, A., 1983, The solubility of calcite and aragonite in seawater at various salinities, temperatures, and one atmosphere total pressure: American Journal of
Science, v. 283, p. 780–799.
NIELSEN, J.K., 2004, Taphonomy in the light of intrinsic shell properties and life
habits: Marine bivalves from the Eemian of northern Russia: Paläeoclimatology
Zeitschrift, v. 78, p. 53–72.
NIELSEN, J.K., and FUNDER, S., 2003, Taphonomy of Eemian marine molluscs and
acorn barnacles from eastern Arkhangelsk region, northern Russia: Palaeogeography, Palaeoclimatology, Palaeoecology, v. 191, p. 139–168.
PANDOLFI, J.M., and GREENSTEIN, B.J., 1997, Taphonomic alteration of reef corals;
effects of reef environment and coral growth form: I, The Great Barrier Reef:
PALAIOS, v. 12, p. 27–42.
PARSONS, K.M., 1993, Taphonomic attributes of mollusks as predictors of environment of deposition in modern carbonate systems: northeastern Caribbean: Unpublished Ph.D. thesis, University of Rochester, Rochester, 434 p.
PARSONS, K.M., and BRETT, C.E., 1991, Taphonomic processes and biases in modern
marine environments: An actualistic perspective on fossil assemblage preservation, in Donovan, S.K., ed., The Processes of Fossilization: New York, Columbia
University Press, p. 22–65.
PARSONS-HUBBARD, K.M., 2005, Molluscan taphofacies in Recent carbonate reef/lagoon systems and their application to sub-fossil samples from reef cores: PALAIOS, v. 20, p. 175–191.
PIP, E., 1988, Differential attrition of molluscan shells in freshwater sediments: Canadian Journal of Earth Sciences, v. 25, p. 68–73.
RICE, W.M., 1989, Analyzing tables of statistical tests: Evolution, v. 43, p. 223–225.
RODLAND, D.L., KOWALEWSKI, M., CARROLL, M., and SIMOES, M.G., 2004, Colonization
of a ‘Lost World’: Encrustation patterns in modern subtropical brachiopod assemblages: PALAIOS, v. 19, p. 381–395.
SANDERS, D., 2003, Syndepositional dissolution of calcium carbonate in neritic carbonate environments: Geological recognition, processes, potential significance:
Journal of African Earth Sciences and the Middle East, v. 36, p. 99–134.
SIMON, A., and POULICEK, M., 1990, Biodegradation anaerobique des structures skeletiques en milieu marin: I. Approche morphologique: Cahiers de Biologie Marine,
v. 31, p. 95–105.
SPEYER, S.E., and BRETT, C.E., 1986, Trilobite taphonomy and Middle Devonian
taphofacies: PALAIOS, v. 1, p. 312–327.
STAFF, G., and POWELL, E.N., 1990a, Taphonomic signature and the imprint of taphonomic history: Discriminating between taphofacies of the inner continental shelf
and a microtidal inlet, in Miller, William III, ed., Paleocommunity temporal dynamics; patterns and processes of long-term community development: Paleontological Society Special Publication, v. 5, p. 370–390.
STAFF, G.M., and POWELL, E.N., 1990b, Local variability of taphonomic attributes in
a parautochthonous assemblage: Can taphonomic signature distinguish a heterogeneous environment?: Journal of Paleontology, v. 64, p. 648–658.
TAYLOR, J.D., KENNEDY, W.J., and HALL, A., 1969, The shell structure and mineralogy
of the Bivalvia. Introduction, Nuculacea-Trigonacea: Bulletin of the British Museum (Natural History), Zoology Supplement, v. 3, 125 p.
TAYLOR, J.D., KENNEDY, W.J., and HALL, A., 1973, The shell structure and mineralogy
of the Bivalvia: II. Lucinacea-Clavagellacea, Conclusion: Bulletin of the British
Museum (Natural History), Zoology, v. 22, p. 253–284.
VERSAR, 2002, Chesapeake Bay water quality monitoring program: Long-term benthic monitoring and assessment component quality assurance plan 2001–2002:
Columbia, Maryland, VERSAR, Inc., 39 p.
WALTER, L.M., 1985, Relative reactivity of skeletal carbonates during dissolution:
Implications for diagenesis, in Schneidermann, Nahum, and Harris, Paul M., eds.,
Carbonate cements revisited: SEPM Special Publication, v. 36, p. 3–16.
WALTER, L.M., and MORSE, J.W., 1984, A re-evaluation of magnesian calcite stabilities: Geochimica et Cosmochimica Acta, v. 48, p. 1,059–1,069.
WALTER, L.M., and MORSE, J.W., 1985, The dissolution kinetics of shallow marine
carbonates in seawater: A laboratory study: Geochimica et Cosmochimica Acta,
v. 49, p. 1,503–1,513.
ZENETOS, A., 1990, Discrimination of autochthonous vs. allochthonous assemblages
in the Eden Estuary, Fife, Scotland, U.K.: Estuarine, Coastal and Shelf Science,
v. 30, p. 525–540.
ZENETOS, A., 1991, Re-evaluation of numerical classification methods for delimiting
biofacies and biotopes in an estuarine environment: Lethaia, v. 24, p. 13–26.
ZUSCHIN, M., and STANTON, R.J.J., 2001, Experimental measurement of shell strength
and its taphonomic interpretation: PALAIOS, v. 16, p. 161–170.
ACCEPTED MARCH 13, 2006